Model studies of terpene biosynthesis. Stereospecific cyclization of N

Juan A. Faraldos , Paul E. O'Maille , Nikki Dellas , Joseph P. Noel and Robert ... Rodney B. Croteau , John J. Shaskus , Britta Renstrom , N. Mark Fel...
2 downloads 0 Views 358KB Size
J . A m . Chem. SOC.1982, 104, 1420-1422

1420

70 (25), 69 (61), 68 (34), 67 (21), 54 (14), 53 (14), 43 (loo), 42 (35), 41 (32), and 39 (16); GC-MS-SIM m / z (integration) 72 (0.1584 X lo5), 71 (0.1911 X lo4); incorporation of ’HI, 94.8%. [ 1,3-2Hz]2-(2’-Propenyl)-5-methyl-4-hexenyl Acetate, [1,3-2H2]Lavanduly1 Acetate ([1,3-2H2]2-OAc). To a vigorously stirred solution containing 0.33 g (2.56 mmol) of ( S ) - [l-2H]3-methyl-2-butenyl acetate in 0.66 g of ethyl acetate at 0 “ C was added 0.17 g (1.28 mmol) of anhydrous aluminum chloride. The reaction mixture was allowed to warm to room temperature and stirring was continued for 110 min before 2 mL of saturated brine was added. The resulting solution was extracted with pentane, and the combined organic layers were dried over anhydrous magnesium sulfate. Solvent was removed at reduced pressure, giving 76 mg (30%) of a colorless liquid. Analytical samples were purified by GLC (Carbowax 20 M, 140 “C); IR (CCI4) 3060, 2955, 2905, 2850, 2310, 1737, 1640, 1445, 1372, 1240, 1100, 1045,900,825 cm-’; N M R (CDCI,) 1.58 (3, s, methyl), 1.68 (6, s, two methyls), 2.00 (3, s, acetoxy methyl), 2.07 ( I , d of d, ’ J = 9 Hz, ,J = 6 Hz, H at C(3)), 2.34 (1, d of d, ’J = 9 Hz, ’J = 6 Hz, H at C(2)), 3.99 (1, br d, ’J = 6 Hz, H at C(l)), 4.70 (1, br s, olefinic methylene), 4.78 (1, br s, olefinic methylene), 5.02 (1, br d, 3J = 6 Hz, H at C(4)); mass spectrum (70 eV) m / z (relative intensity), 139 (2). 138 (S), 123 (13), 95 (47), 84 (22), 83 ( I O ) , 81 (14), 75 (15), 70 (loo), 69 (57), and 68 (23); GC-MS-SIM (CI, isobutane) m / z (integration) 199 (0.1422 X lo’, M I ) , 198 (0.8298 X lo3), 197 (0.1968 X lo’); incorporation of ’H,, 93.4%. [2,4-2H2]3-(2’-PropyI)butyrolactone ([2,4-2H2]3). Following the procedures described previously for preparation of 3-(2’-propyl)butyrolactone from lavandulyl acetate, 76 mg (0.38 mmol) of [ 1,3-2H2]lavandulyl acetate ([ 1,3-2H2]2-OAc)was converted into [2,4-2H2]3-(2’-propyI)bu-

+

trolacetone ([2,4-’H2]3). The crude product was purifed by GLC (Carbowax 20 M, 140 “C) to yield 15.4 mg (31%) of a colorless oil; N M R (C6D6, 300 MHz), 0.36 (3, d, = 6.6 Hz, methyl), 0.41 (3, d, ’ J = 6.6 Hz, methyl), 0.82 (1, d of heptets, ’ J = 6.6 Hz, ’ J = 6.6 Hz, ’J = 8.4 Hz, methine H of isopropyl), 1.30 (1, br m, ’ J = 8.4 Hz, ,J = 9.9 Hz, ’ J = 8.6 Hz, ’ J = 7.6 Hz, H at C(3)), 1.52 (0.5, d o f t , ’J = 9.9 Hz, ’J = 2.5 Hz, H at C(2) trans to H at C(3)), 1.92 (0.5, d o f t , 3J = 8.6 Hz, 2J = 2.5 Hz, H at C(2) cis to H at C(3)), 3.20 (0.5, br d, ’ J = 8.6 Hz, H at C(4) trans to H at C(3)). 3.66 (0.5, br d, 3J = 7.6 Hz, H at C(4) cis to H at C(3)).

Acknowledgment. We wish to thank Professor W. W. Epstein for a generous gift of (S)-(+)-2,2,2-trifluoro- 1-(9’-anthry1)ethanol.

Registry No. I-OAc, 1191-16-8; (lS)-[1-2H]I-H, 55833-58-4; (lS)-[ l-’H]l-OAc, 80410-17-9; 2-OAc, 25905-14-0; (lS,2R,3R)-[ 1,3’H2]2-OAc, 80410-18-0; (1S,2R,3S)-[1,3-’H2]2-OAc, 80446-34-0; ( ~ S , ~ S , ~ R ) - [ I , ~ -80446-35-1; ’ H ~ ] ~ - O(~S,ZS,~S)-[I,~-’H~]~-OAC AC, 80446-36-2; (R)-3, 80410-19-1; ( 9 - 3 , 53657-15-1; (k)-3, 80446-37-3; (2S,3S,4S)-[2,4-2H2]3, 80410-20-4; (2S,3R,4S)-[2,4-’H2]3, 80446-38-4; (2R,3S,4S)-[2,4-’H2]3, 80446-39-5; (2R,3R,4S)-[2,4-’HJ3, 80446-40-8; 4-OAc, 74912-37-1; 5-OAc, 80410-21-5; (R)-7, 4221-98-1; (R)-8, 1187-69-5; ( R ) - 9 , 80410-22-6; (R)-10, 80410-23-7; 12, 109-92-2; 13, 107-86-8; [l-’H]13, 21849-61-6; 14, 19860-69-6; [2-’H]14, 80410-24-8; (lS)-[1-’H] 15, 55833-58-4; tetrahydrolavandulyl acetate, 40853-55-2; 3-methyl-2-butenyl acetate, 3814-41-3; 2-(2’-methylpropen-I’-yl)-1,3dithiane, 19860-69-6; [ l-’H]isoamyl alcohol, 53939-07-4.

Communications to the Editor Model Studies of Terpene Biosynthesis. Stereospecific Cyclization of N-Methyl-(S)-4-([1’-*H]neryloxy)pyridinium Methyl Sulfate to a-Terpineol C. D a l e Poulter* a n d Chi-Hsin R. King

-I-x

Department of Chemistry, University of Utah Salt Lake City, Utah 84112 Receiued August 10, I981

Much of t h e structural diversity found in t h e terpene biosynthetic pathway is introduced by olefin cyclization reactions of a few common acyclic precursors. The basic strategy involves intramolecular electrophilic alkylation of remote double bonds by allylic pyrophosphate esters. Treatment of the acyclic terpenes nerol ( 1 - O H ) and linalool ( 3 - O H ) with acid or solvolysis of appropriate derivatives gives a complex mixture of products containing 1 - O H , 3-OH, geraniol (2-OH), a n d a-terpineol ( 4 - O H ) . Formation of t h e cyclic isomer from neryl and linalyl precursors is c o m m o n l y assumed t o be a good model for related biological direct a n d allylic displacements.’ As early as 1898, S t e p h a n 2 reported optical induction in t h e acid-catalyzed cyclization of 3-OH t o 4-OH, a n d enantiomeric excesses of up t o 90% were found during the cyclization of linalyl p-nitrobenzoate ( ~ - O P N B ) . ~ Arigoni and c o - ~ o r k e r sin , ~a particularly elegant piece of work, recently deduced which conformation of t h e linalyl skeleton was preferred during cyclization. Several groups have commented on t h e attractiveness of a concerted reaction with P participation by t h e remote double bond t o explain t h e stereoselectivity observed for the allylic d i ~ p l a c e m e n t . Although ~~ the isomeric neryl system has been studied e ~ t e n s i v e l y , ~ there ~ ~ - ’ ~a r e no reports of stereo( 1 ) See Cane (Cane, D. E. Tetrahedron 1980, 36, 1109-1 159) for a definition of direct and allylic displacements. (2) Stephan, W. J. Prakt. Chem. 1898, 58, 109. (3) Winstein, S.; Valkanas, G.; Wilcox, C. F. J . A m . Chem. Soc. 1972, 94, 2286-2290. (4) Godtfredsen, S.; Obrecht, J. P.; Arigoni, D. Chimia 1977, 31, 62-63. ( 5 ) Rittersdorf, W.; Cramer, F. Tetrahedron 1968, 24, 43-52.

0002-7863/82/ 1504-1420$01.25/0

+$

HQ

I-OH

2-oH

.f

:-OH

OH

4-OH

X = halogen, hydroxyl, p-nitrobenzoate, phosphate, pyrophosphate chemical studies for cyclization of 1-OH t o 4-OH, presumably because of difficulties associated with the lack of a chiral center (6) Coates, R. M. Prog. Chem. Org. Nut. Prod. 1976, 33, 74-230. (7) Kitagawa Y.; Hashimoto, S.;Satoshi, I.; Yamamoto, H.; Nozaki, H. J . A m . Chem. SOC.1976, 98, 5030-5031. (8) Bunton, C. A.; Cori, 0.;Hanchey D.; Leresche, J. P. J. Org. Chem. 1979, 44, 3238-3244. (9) Rittersdorf, W.; Cramer, F. Tetrahedron 1967, 23, 3015-3022. (10) Ritersdorf, W.; Cramer, F. Tetrahedron 1967, 23, 3023-3028. ( 1 1) Valenzuela, P.; Cori, 0. Tetrahedron Lett. 1967, 3089-3094. (12) Bunton, C. A,; Leresche, J. P.; Hanchey, D. Tetrahedron Lett. 1972, 243 1-2434.

0 1982 American Chemical Society

J. Am. Chem. Sot., Vol. 104, No. 5 , 1982 1421

Communications to the Editor Scheme I. Synthesis of /V-?rlethyl-(S)-4-([ l'-'H] nery1oxy)pyridinium Methyl Sulfate

A+o%

y

m

C H30

5

,

9 , 'H

H OH

[ 111 -*ti']

L-OH

Scheme 11. Conversion of a-Terpineol to 3 4 3'-Oxo bu tyl)-4,4-dime t hyl but ry olact one

[I- 'ti]

0

6

1.

yI=3,, a

;I

OPY+

H

MeSO; (s)-[l-'H]~-OPy+MeSO~

MnO,, CH,CI,, 25 h. LiAl'H,, Et,O, - 1 8 'C, 1 2 h. KH, T H F , 1 h. e 4-Chloropyridine, 1 2 h. Yeast, sucrose. Dimethyl sulfate, Et,O, 4 h.

'

O,, --78"C.

in the molecule. In this communication we discuss experiments with chiral, isotopically labeled 1-OH which permit us to deduce the stereochemistry a t C ( l ) during a direct intramolecular displacement, and in the following communication we address the timing of ionization and c y ~ l i z a t i o n . ' ~ Our conclusions have important mechanistic implications for electrophilic alkylations of remote double bonds by allylic moieties in biomimetic and enzyme-catalyzed reactions. The stereochemistry a t C ( 1) during the direct displacement represented by the neryl to a-terpinyl cyclization was studied by using N-methyl-(S)-4-( [ 1'-2H]neryloxy)pyridinium methyl sulfate ((S)-[ l'-2H] I-OPy+MeSO,-), a chiral derivative with a neutral leaving g r 0 ~ p . l ~The synthesis of ( S ) -[ 1'-2H] l-OPy'MeS0,shown in Scheme I took advantage of the known stereoselectivity of dehydrogenase activity in actively fermenting yeast's-'7 to prepare (S)-[l-2H]nerol ( ( S ) - [1-2H]1-OH)'s from labeled aldehyde [l-2H]6. Hydrolysis of the pyridinium salt in 0.14 M aqueous sodium bicarbonate a t 25 "C gave a mixture of labeled 1-OH (l%), 3-OH (15%), and 4-OH (82%), along with some unidentified hydrocarbons (2%).19 The two major components were purified by GL.PC on Carbowax 20M. Cyclization of (S)-[l'-ZH]1-OPy+MeSO4-across C ( 1) and C(6) gives [3-*H]4-OHwith chiral centers at C(3) and C(4). Assuming that deuterium does not generate any measurable amount of optical induction during cyclization, C(4) is formed stereorandomly. If cyclization is stereospecific at C ( l), [3-2H]4-OH consists of a mixture of two stereoisomers, 3R,4R and 3R,4S for inversion or 3S,4R and 3S,4S for retention. If, however, cyclization is stereoselective or stereorandom, [ 3-2H]4-OH consists of a mixture of all four diastereomers. In the latter case, the selectivity of the reaction can be deduced from the relative proportions of the stereoisomers. The composition of the stereoisomeric mixture was determined by 'HN M R spectroscopy after conversion of labeled a-terpineol to lactone 7 as shown in Scheme 11. This sequence translates C(3) and C(4) of 4-OH into C(2) and C(3) of 7 without changing the configuration of the chiral centers. Chemical shift assignments of the protons at C(2) and C(3) were made by using unlabeled lactone prepared from racemic a-terpineol and are listed in Table 1. A f o u r - h e pattern a t 2.05 ppm was assigned to Hb on the (13) Poulter, C. D.; King, C. H. R. J . Am. Chem. SOC.1982, 104, 1422-1426. ( 1 4) This derivative was selected in order to minimize stereorandomization of the label by ionization and internal return. (15) Althouse, V. E.; Feigl, D. M.; Sanderson, W. A,; Mosher, H. S. J . Am. Chem. So'. 1966, 88, 3595-3599. (16) Poulter, C. D.; Hughes, J. M. J . Am. Chem. SOC.1977, 99, 3824-3829. (17) Gerlach, H.; Zagalak, B. J . Chem. Soc., Chem. Commun. 1973, 214-27 5 . (18) All labeled compounds gave satisfactory IR and NMR spectra. Incorporation of deuterium in (S)-[1-2H]l-OHwas greater than 9996, as determined by mass spectrometry. (19) The absence of trace quantities of geraniol (2-OH), commonly found with anionic leaving groups, indicates that internal return had been suppressed.

Jones reagent.

' C,H,,

A.

Table I. NMR Parameters for Ha, Hb, and Hc in 3-(3'-0xobuty1)4,4-dimethylbutyrolactone chemical shifts

(S)-[I-*H]L-OH

a

-7

-OH

compd

dR

- Ss,

da

SxR

62q

Hz

Ha

1.69

l.llb

1.22b

-33.0

Hb

2.05 1.49

1.57b

1.5Sb

Ha

1.66

d

1.23'

Hb

2.01 1.49

1.57'

1.55'

coupling constants, Hz

7

HC [ 2-, H ] 7

HC

6.4

, J H , , H ~ = 17.0 3 J ~ a , =~ 11.6 , 3 Jb*H~c = 8.0

6.2

, J H , , H ~ = 2.4 3 J ~ a ,=~11.6 c ' J H ~ ~=H 8.0,

a Recorded at 3 0 0 MHz, 84 mM solution in benzene-d, with Me,Si internal standard. Measured for 7 using a 90 mM solution of a mixture containing 1 part (R)-7, 2 parts (R,S)-7, and 7.2 equiv of (S)-2,2,2-trifluoro-l-(9'-anthryl)ethanol. Measured for [ 2-,H] 7 using a 74 mM solution of a mixture containing 1 part of [2-,H]7 from ( S ) - [l'-'H] 1-OPy'MeSO;, 1.5 parts of [2-'H]7 from (R,S)-[]'-,HI l-OPy+MeSO,-, and 11.2 equiv o f (S)-2,2,2trifluoro-l-(9'-anthryl)ethanol. Obscured by signals from the butyl side chain between 1.05-1.15 ppm.

'

'

basis of an 8 f 2% NOE enhancement when a complex multiplet a t 1.49 ppm, assigned to H,, was irradiated. Upon addition of 7.2 equiv of Pirkle's chiral shift reagent,20 (S)-2,2,2-trifluorol-(Y-anthryl)ethanol, to a solution of 7 in benzene-d6, the four-line patterns at 1.69 and 2.05 ppm assigned to Ha and Hb, respectively, each moved upfield and separated into two well-resolved sets. In a separate experiment (R)-7, prepared from (R)-4-OH,21-23was mixed with racemic lactone, and the 'H N M R spectrum was recorded in the presence of the chiral shift reagent. Enhanced intensities were seen for peaks in the upfield pattern for H a and in the downfield pattern for Hb. Thus, we could measure signals from Ha and Hb in (R)-7 and in (54-7 directly without separating the stereoisomeric lactones. [3-2H]a-Terpineol obtained from solvolysis of (S)-[l'-ZH] 1OPy+MeSO,- was then converted to [2-2H]7 for the N M R measurements. The signals for Hb are shown in Figure l a . The six-line pattern a t 2.01 ppm is characteristic of a large vicinal coupling to H, (3JH,,H, = 8.0 Hz) and a small geminal coupling to the deuteron ('JH,,H, = 2.4 Hz). Addition of 7.2 equiv of chiral shift reagent to the solution shifted the pattern upfield to 1.57 ppm, along with some loss of resolution. It is apparent, however, from examination of the spectrum shown in Figure l b that the resonance for H b did not separate into enantiotopic pairs as we had previously observed with racemic, unlabeled lactone. Finally, 1.5 equiv of labeled lactone prepared from [3-2H]4-OH obtained by cyclization of (R,S)-[l'-2H] l-OPy+MeSO,- were added to the sample. The broad doublet a t 1.57 ppm became the three-line (20) Pirkle, W. H.; Sikkenga, D. L.; Pavlin, M . S. J . Org. Chem. 1977, 42, 384-387. (21) (R)-4-OH was obtained by oxymercuration of ( R ) - l i m ~ n e n e ~ ~ , ~ ~ [Aldrich, [a]240f106' (c I , EtOH)]. (22) Brown, H. C.; Geoghegan, P. J.; Lynch, G.J.; Kurek, J. T. J . Org. Chem. 1972, 37, 1941-1947. (23) Bambagiotti, A. M.; Vincieri, F. F.; Coran, S. A. J . Org. Chem. 1974, 39, 680-682.

1422

J . Am. Chem. SOC. 1982, 104, 1422-1424

preferentially from an anti-endo c ~ n f o r m a t i o n ,and ~ a similar preference is expected for its allylic isomer. The stereoselectivity

a

-

-

+ Me

+

pyoM pyoLd S04.Me

anti - exo

anti-endo

we observed a t C ( l ) for the direct process is measurably higher than the preference reported for the anti-endo mode in the allylic displacement. The difference might indicate that allylic cyclization can also occur by competing anti-exo or syn modes. The former possibility cannot be detected by the technique we employed, and the latter is precluded for a direct displacement. Linalyl pnitrobenzoate was, however, used to study the allylic displacement: and loss of stereocontrol could have resulted from internal return of the anionic leaving group. As mentioned earlier, a concerted mechanism offers an attractive rationale for the stereochemistry of direct and allylic displacements. It must be emphasized, however, that while a concerted electrophilic cyclization requires stereospecificity, the converse-that stereospecificity establishes concertedness-does not hold. A stepwise process where cyclization is faster than reorientation of the side chain is also consistent with the stereochemistry of the electrophilic cyclizations. This question is addressed in the following communication.

b

C

Acknowledgment. This work was supported by the Institute of General Medical Science of the National Institutes of Health, G M 21328. Registry No. l-OPy+MeSO,-, 80387-02-6; a-terpineol, 98-55-5; (3R,4R)-[3-2H]-4, 80375-27-5; (3R,4S)-[3-2H]-4, 80408-85-1; (2S,3R)-[2-2H]-7,80375-28-6; (2S,3S)-[2-2H]-7, 80408-86-2; (R)-7, 38746-47-3; ( 9 - 7 , 80408-87-3; ( ( S ) - [1’-2H]l-OPy*MeSO,-), 8038704-8.

Model Studies of Terpene Biosynthesis. A Stepwise Mechanism for Cyclization of Nerol to a-Terpineol

10 Hz

Figure 1. ’H NMR spectra for Hbin [2-2H]3-(3’-oxobutyl)-4,4-dimethylbutyrolactone] ([2-2H]7)recorded in benzene-d, at 300 MHz. (a) [2-2H]7from [3-*H]4-OHobtained during solvolysis of (S)-[1’-2H]1OPy+MeS04-; (b) same as (a) with 7.2 equiv of (S)-2,2,2-trifluoro-l(9’-anthryl)ethanol added; (c) same as (a) with 1.5 equiv of [2-2H]7 obtained from (R,S)-[1-2H]1-OH and 1 1.2 equiv of (S)-2,2,2-trifluoro1-(9’-anthryl)ethanol added.

pattern shown in Figure IC, formed from overlapping doublets ( J = 8 H z j with the low field pattern being more intense. Similar behavior was seen for Ha. According to the assignments presented in Table I, the deuteron in [2-2H]7 is cis to H, when C(3) is S and trans when C(3) is R. These experiments clearly establish that [2-2H]7 obtained from (S)-[1-2H]1-OH is a mixture of only the 2S,3R and 2S,3S diastereomer^.^^ Labeled a-terpineol derived from (S)-[ l’-2H] 1-OPy+MeSO,- must, therefore, consist of only the 3R,4R and 3R,4S stereoisomers. The obvious conclusion is that cyclization is stereospecific a t C(l) and proceeds with inversion of configuration. It follows that the fraction of l-OPy+MeSO,- which cyclizes to a-terpineol must d o so from a conformation where the remote double bond is positioned a t the backside of C( 1). Two limiting orientations which accomodate this restriction are shown below. Although we cannot distinguish between the anti-endo and anti-exo modes, the topologically related linalyl system is known to cyclize (24) We saw no peaks for the 2R,3R and 2R,3S stereoisomers, even at high spectrum amplitudes, and conservatively estimate that they constituted less than 5% of the mixture.

0002-7863/82/ 1504-1422$01.25/0

C. Dale Poulter* and Chi-Hsin R. King Department of Chemistry, University of Utah S a l t Lake City, Utah 84112 Received August I O , 1981 Electrophilic alkylations of remote double bonds by allylic moieties are important carbon-carbon bond forming reactions in terpene metabolism and related biomimetic olefin cyclizations.’-4 The enzymatic and nonenzymatic reactions are both characterized by a high degree of stereoselectivity. Two explanations have evolved for this p h e n o m e n ~ n . ~One is the reactions are concerted. This is attractive since stereospecificity is a logical result of the synchronous changes in bonding that occur in concerted reactions. The other explanation is a nonconcerted process involving a series of conformationally rigid intermediates where topology is maintained between the initiation and termination step^.^,^,^-'^ ( 1 ) Coates, R. M. Prog. Chem. Org. Nut. Prod. 1976, 33, 74-230. (2) Poulter, C. D.; Rilling, H. C. Acc. Chem. Res. 1978, 11, 307-313. (3) Cane, D. E. Tetrahedron 1980, 36, 1109-1 159. (4) Poulter, C. D.; Rilling, H. C. “Biosynthesis of Isoprenoid Compounds”; Porter, J . W., Ed., Wiley: New York, 1981; Vol. 1, pp 161-224. (5) Johnson, W. S . Angew. Chem., Int. E d . Engl. 1976, 15, 9-17. ( 6 ) Poulter, C. D.; Wiggins, P. L.; Le, A. T. J . Am. Chem. SOC.1981, 103, 3926-3927. (7) Mash, E. A,; Gurria, G. M.; Poulter, C. D. J . Am. Chem. SOC.1981, 103, 3927-3929.

0 1982 American Chemical Society